抄録
This paper describes 5-axis centralized feedback control system for a single-rotor helicopter. Optimal control theory is directly applied to 25-state Multiple-Input Multiple-Output (MIMO) analytical linear model of a small-scale unmanned helicopter without decoupling the motion dynamics. The linear Kalman filter is also designed for an estimation of the blade flapping and lagging angles and the down-wash which are utilized in the state feedback control law. In the comparison with the distributed control system using the Single-Input Single-Output (SISO) controller which is common approach for flight control, this full-axis centralized control system includes the effects of cross-coupling dynamics and enables a design for full control-axis at once in a short amount of time. Flight test with this method demonstrated a steady hover control performance through attitude, position, and main rotor speed. Simple tuning way of weighting matrix used in the quadratic cost function is also presented to deal with a difficult problem caused by a large number of design parameters or the undesirable coupled vibration. Stability analysis of the closed loop system using MIMO plant model reveals that MIMO controller provides greater and better balance stability margins for every control-axis than SISO controller does. In addition, a frequency characteristic analysis is conducted using MIMO plant models based on different flight conditions, which showed a good robustness of the MIMO controller toward modeling errors and also the effect of extra flight conditions on the robustness of each control-axis.
